A Proximity Determinable Social Pseudonym Framework in Online Identity Management System

Online identities play critical roles in the current Internet world. One of the virtues of traditional online identities, in forms of pseudonyms, is the privacy protection of online users. Users can create as many anonymous identities as they want with only an email account. However, the anonymity itself could be a limitation because of the lack of accountability.;The prevalence of online social networks inspires us to create a social pseudonym framework through which it is possible to determine proximity between pseudonyms while retaining privacy. The basic idea of this work is mapping the online social network into a geometric space and assigning each user a coordinate according to the relationship with his/her neighborhood. The coordinate itself will not disclose any information of a user, but by computing the coordinates distance, proximity between users could be estimated with a certain probability. I evaluated the framework with several real online social network datasets. The results indicate that the proposed framework is promising under certain conditions.